The combined effect of Pdx1 overexpression and Shh manipulation on the function of insulin-producing cells derived from adipose-tissue stem cells

Abstract

Pancreatic and duodenal homeobox 1 (Pdx1) and Sonic hedgehog (Shh) are the key regulators of beta-cell function. In vitro experiments have shown that there is significant cooperation between Pdx1 and Shh with regard to the production and maintenance of insulin-producing cells (IPCs). In this study, the combined effect of Pdx1 overexpression and Shh manipulation on the function of adipose tissue-derived IPCs was determined. A eukaryotic expression vector (Pdx1- pCDNA3.1(+)) was constructed and transfected into a Chinese hamster ovary (CHO) cell line. Adipose tissue-derived mesenchymal stem cells (ADMSCs) obtained from rats were assigned to two groups [control (C) and manipulated (M)] and differentiated into IPCs. Manipulated cells were treated with a mixture of FGF-β and cyclopamine and recombinant Shh protein at days 3 and 11, respectively, and transfected with Pdx1- pCDNA3.1(+) at day 10. The expression of multiple genes related to function of beta cells was analyzed using real-time PCR. The functionality of IPCs in vitro was analyzed through dithizone (DTZ) staining and ELISA. IPCs were injected into the tail vein of diabetic rats, and blood glucose and insulin concentrations were measured. CHO cells transfected with Pdx1- pCDNA3.1(+) showed a significantly higher expression of Pdx1 compared with nontransfected cells. Manipulated IPCs exhibited a significantly higher expression of MafA, Nkx2.2, Nkx6.1, Ngn3, insulin, and Isl1 and a higher insulin secretion in response to glucose challenge in relation to control cells. Rats that received manipulated IPCs exhibited a higher ability to normalize blood glucose and insulin secretion when compared to controls. Our protocol might be used for more efficient cell therapy of patients with diabetes in the future.

Keywords: Pdx1; adipose tissue‐derived mesenchymal stem cells; insulin‐producing cells; sonic hedgehog pathway.

Figure 1

Differentiation protocol of ADMSCs into IPCs. ADMSCs in the control group were differentiated into IPCs using the three‐stage basic protocol. Cells in the manipulated group were treated with 0.25 μm cyclopamine and 64 ng·mL⁻¹ FGF‐β at day 3 of differentiation for suppression of Shh pathway and were subsequently treated with recombinant Shh at day 11 of differentiation for reactivation of Shh pathway. Manipulated cells were transfected with Pdx1‐ pCDNA3.1(+) recombinant plasmid at day 10 of differentiation.

Figure 2

(A) Characterization of Pdx1‐ pCDNA3.1(+) vector. Lane 1: the recombinant plasmid before digestion. Lane 2: the 850‐bp Pdx1 gene separated from the recombinant Pdx1‐ pCDNA3.1(+) digested using EcoRI and HindIII enzymes. Lane 3: 900‐bp PCR product of recombinant Pdx1‐ pCDNA3.1(+) vector. Lane 4: 1‐kb DNA ladder. (B) Western blot analysis results. The CHO cells transfected with Pdx1‐ pCDNA3.1(+) (lane 1) showed a significantly higher expression of Pdx1 compared with the CHO cells transfected with pCDNA3.1(+) alone (Lane 2).

Figure 3

Characterization of ADMSCs. Flow cytometry analysis of CDs markers in isolated ADMSCs showed that undifferentiated ADMSCs were positive for MSC‐specific surface marker proteins including CD90 (C,G) and CD105 (D,H) and negative for hematopoietic and leukocyte common antigens including CD34 (A,E) and CD45 (B,F). The presence of Oil Red O‐stained oil droplets confirmed the adipogenic differentiation potential of ADMSCs (I). Osteogenic differentiation capacity of ADMSCs was confirmed by formation of Alizarin Red‐stained calcium phosphate deposits (J). Phase contrast magnification, 100×.

Figure 4

Morphological characteristics of ADMSCs during the differentiation into IPCs. Spindle‐like ADMSCs were shortened and aggregated at day 3 of differentiation (A). Cells presented an epithelial‐like morphology at day 7 of differentiation (B). Cells exhibited an epithelial cell morphology at day 14 of differentiation (C). IPCs stained as crimson red by DTZ at late stages of differentiation (D). Phase contrast magnification, 100×.

Figure 5

Expression of pancreas‐related genes after Shh pathway manipulation and Pdx1 overexpression: Manipulation of Shh caused an increase in Pdx1, MafA, Nkx2.2, Nkx6.1, Ngn3, Isl1, and insulin expression compared with the control group (P < 0.05). GAPDH was used as the calibrator for real‐time PCR analysis. Data are expressed as mean ± SE. Statistical significance difference at P < 0.05 is represented by different letters.

Figure 6

The results of insulin secretion assay. Undifferentiated ADMSCs showed no obvious ability for insulin secretion in response to glucose (A,B). The manipulated group showed a higher expression and secretion of insulin in comparison with the control group in vitro (P < 0.05) (A). The rats that received manipulated IPCs secreted significantly higher amounts of insulin in response to glucose when compared with the rats that received control IPCs (P < 0.05) (B). Statistical significance difference at P < 0.05 is represented by different letters.

Figure 7

Fasting blood glucose in rats treated with different IPCs. After transplantation of manipulated IPCs to STZ‐diabetic rats (n = 5), a sharp reduction in the mean blood glucose concentration (92 ± 1.2 mg·mL⁻¹) within 2 weeks was noticed. At the end of the third week, the average amount of blood glucose concentration raised to 315 ± 1.9 mg·dL⁻¹. Thereafter, blood glucose concentration decreased gradually. At the sixth week after transplantation, the average amount of glucose concentration reached 147 ± 1.1 mg·mL⁻¹. The rats that received positive control IPCs (n = 5) followed the same pattern of blood glucose concentration changes. However, they showed a more severe hyperglycemic condition. The diabetic rats that received undifferentiated ADMSCs showed no detectable change in the mean blood glucose concentration.